Sensor-Based Irrigation: When to Water and How to Verify Delivery

Define a zone that can share one irrigation decision

Group plants by crop, growth stage, container, substrate and hydraulic supply. Identify sunny edges, different pot sizes and other conditions that may need a separate zone. Write down the acceptable root-zone moisture range with the crop manager and distinguish routine triggers from emergency limits. A clock-based programme is a useful baseline to measure against. The proposed upgrade should answer a clear question, such as whether watering stays appropriate as the canopy grows or whether failed delivery is detected promptly.

Make the moisture reading meaningful

Volumetric water content expresses the fraction of substrate volume occupied by water. It is not the same as the fraction of plant-available water. Check the manufacturer’s calibration for the actual medium, its salinity range and temperature behaviour. Install probes with consistent contact, depth and position relative to the emitter. Air gaps, disturbed media and a probe in a locally saturated pocket can distort the decision. Use several representative points during commissioning and compare with an independent agreed reference rather than trusting one convenient pot.

Use bounded pulses with time to observe the result

Nemali and van Iersel’s 2006 experiment demonstrated moisture-triggered irrigation with calibrated sensors and valves in potted plants. Original experiment. It supports the feedback principle, not copying its set-points into another substrate or crop. Hamfy’s proposed sequence is: validate the reading, confirm irrigation is permitted, deliver a bounded pulse, then wait an agreed redistribution interval before reassessing. Separate that interval from the maximum allowed delay when the crop is genuinely too dry. Keep a daily volume limit and a defined operator escalation.

The diagram below uses invented moisture readings to show why the immediate post-pulse value and the later settled value differ. It is not a measured crop response or a recommended threshold. During a real trial, retain raw readings, command time, pulse duration and delivered volume; otherwise a trend cannot distinguish plant demand from a change in the irrigation programme.

Check the hydraulic system as well as the software

UGA’s startup checklist pairs system maintenance with pressure and flow checks. Delivery checks. Record expected zone flow at an operating pressure, then test representative emitters near the beginning and end of laterals. Consider filter condition, line filling, leaks and drainage between pulses. A normal total flow does not prove uniform delivery to every container. Keep water quality and deposits in the investigation through our pH and EC guide. A software dashboard cannot compensate for an unbalanced or blocked distribution system.

Use a volume check to challenge the command

Worked example with assumed steady flow: 100 emitters at 2 L/h running for 3 minutes would deliver 10 L in total, or 0.1 L per emitter. Compare this expectation with the zone meter after allowing for its resolution and the system’s filling and draining behaviour. A measured 6 L is a discrepancy to investigate, not proof of a particular fault. Check pressure, actual run time, blockage and meter operation before changing the moisture target. Short pulses make those transient effects especially important.

Design responses for missing or conflicting evidence

If the command is on but flow is absent, check the pump, supply, valve and instrumentation. If total flow looks normal but one location remains dry, check local delivery and probe placement. If flow continues after shutdown, investigate a leak, siphoning or a valve fault. Stale, impossible or frozen sensor readings should not silently request unlimited watering. Define bounded fallback behaviour with the grower, notify the assigned operator and keep a manual recovery procedure. The safest response depends on the crop and equipment; automatic shutdown is not universally sufficient.

Evaluate crop outcomes and water use together

Compare matched zones or comparable cycles using total delivered water, drainage, substrate trends, interventions and marketable output. Record radiation, growth stage and major climate changes so a cloudy week is not mistaken for a controller improvement. If fertigation is involved, include the agreed root-zone and drainage checks rather than pursuing minimum drainage regardless of salt balance. Use the sensor-data analyser to inspect missing values and timing. Hamfy proposes this record as a basis for control improvements; it does not establish a guaranteed saving or replace crop judgement.